audiocommander

nice to hear :) So I think I will order one too and experiment... I found this shop, that has same prices like sander and besides the SpeakJet also other nice stuff (eg the sensors as usual 1/3 cheaper than Conrad 8) ) : http://www.shop.robotikhardware.de/ I found also one bausatz (BSRNSPEAK, not using SpeakJet) that seems to have some nice features (54 EUR). As far as I understood, one can upload wav files via RS232 and trigger these very quickly by I2C (I presume it's the same as IIC?). One should upload phonemes, but you could surely do funny things with it, if you don't want to produce clear speach: http://www.shop.robotikhardware.de/shop/catalog/product_info.php?products_id=21

I'm curios: has any progress been made with these nice SpeakJet SynthChips and MIOS? Is Captain Hastings ("Good Lord!" I love Poirot! :D) still around? There's an exhibition scheduled for 2007 and I'll have to develop a speaking machine. I'm not sure if I will do this just by programming... but this chip and MIOS control seems a nice option if there have been experiences made... anyone?

indeed, seems to be a nice hack 8) I hope for you, that Live5 Live6 has the same options available if released this summer/autumn (when your box is finished) ;D (just joking....) Edit: Sorry, meant Live6

I don't think that LIVE sends a status back by midi. How should it? (Besides the fact that this would generate some nice Feedback Loops) And I also don't think that LIVE is scriptable, and if it were: what would you do? (Besides the fact that scripting is always quite slow) Next problem is that there isn't only ON/OFF in LIVE. There is ON/OFF/LOOP/ONCE/TRIGGER... Question: Why don't you simply look on your monitor? Or buy a touchScreen overlay? Should be a lot easier than soldering 127 * xTracks LEDs/Buttons? What if you had more tracks than there are on your controller? Best, Michael

have you checked out the SongPositionPointer? It seems it's exactly what you're looking for! If Live sends this continuous pointer, you can relatively change it by an ENC and send the updated songPositionPointer back to Live. I haven't tried, but don't see a reason why it shouldn't work You'll find in depth info about the spp here: http://www.borg.com/~jglatt/tech/midispec.htm -> Song Position Pointer

Do you mean the Transport Bar in Live's "horizontal" edit mode (not the vertical Slot Mode)? I don't think you can assign a midi controller to that. Maybe LIVE recognizes standard MIDI Song Position (System Common & System Realtime) Messages: take a look here: http://www.midi.org/about-midi/table2.shtml 11110000= F0= 240| System Exclusive | ** | ** 11110001= F1= 241| MIDI Time Code Qtr. Frame | -see spec- | -see spec- 11110010= F2= 242| Song Position Pointer | LSB | MSB 11110011= F3= 243| Song Select(Song #) | (0-127) | NONE 11110100= F4= 244| Undefined (Reserved) | ? | ? 11110101= F5= 245| Undefined (Reserved) | ? | ? 11110110= F6= 246| Tune request | NONE | NONE 11110111= F7= 247| End of SysEx (EOX) | " | " 11111000= F8= 248| Timing clock | " | " 11111001= F9= 249| Undefined (Reserved) | " | " 11111010= FA= 250| Start | " | " 11111011= FB= 251| Continue | " | " 11111100= FC= 252| Stop | " | " 11111101= FD= 253| Undefined (Reserved) | " | " 11111110= FE= 254| Active Sensing | " | " 11111111= FF= 255| System Reset | " | "

Thank you!! :D Exactly what I needed!

Hi! I have only 1 button on my new "project"-box. Yep. 8) But: Is there a way to connect just one button directly to the core? Soldering a whole DIN board seems a bit of an overkill :-\ First I thought I could connect it directly to the core, but I realized it's not like the 8 unmuxed AIN pins. I don't know what to do with the S0, SC and RC pins... Then I wanted to connect it to J14. I init with "MIOS_SRIO_TS_SensitivitySet(0x00);" as I read in the function reference, but "PORTDbits.RD4" is always FALSE and when I measure I get only ~0 V. Has anyone a suggestion or hint for me? ??? (or at least a refreshing one-liner?) ;D Thanks!

that had to be prevented!! mission complete! ;D no, really: I just moved some days ago to Xcode by myself when starting this debug thing. It would be a shame not sharing this info, 'cause Xcode is a very good tool. The "PATH=/usr/local/bin:/other/paths/bin:$PATH" line has been hard to find, but once it was found, everything is ridicilously easy! A new thread for Xcode related issues sounds good!

here you go 8) http://www.midibox.org/dokuwiki/doku.php?id=how_to_use_xcode2_as_ide_on_a_mac thank you all for your warm words (and karmas) :D

http://www.midibox.org/dokuwiki/doku.php?id=mios_c_simulator_-_debugger I hope that this will be a start ;)

As mentioned in another thread, I have written a debug.c-file which helps me to debug the whole app simulating MIOS I/O with standard console input, eg. typing: "a 1,420" would result in a call to "AIN_NotifyChange(pin 1, value 420)". Also, I've wrapped the LCD-Outputs to go straight to the console --eg: extern void MIOS_LCD_PrintBCD3(unsigned char v) { printf("%3.3i",v); } However I don't have the time to do it nicely and implement it _all_ ... would the Wiki be a place for this (everyone could improve and add functions, because I only implemented those I needed!)? But I don't want to shock any newbies...and it's really in a very, very early state... any thoughts from a Guru?

hmm... but this is real fun 8) :

(ok, I see you answered it by yourself... nevertheless here is what I've written for you:) With the C-Skeleton comes a Makefile Generator (located in tools "mkmk.pl"), so you just have to config the "MAKEFILE_SPECS", then generate your Makefile (System Dependent) with mkmk.pl and then run makefile.bat or call "make" in terminal. That means you can use any IDE you want to. Most IDEs support simple calls to makefiles. Personally I'm using Xcode2 on Mac OS X doing exactly what you describe, pressing "Build" and getting the .syx-file generated :) Michael.

yeah! I think root is the right word: damn translation widgets! ::) I'll exchange the signs :) Thanks! There's also one point about this method: I had real problems without these load resistors, even with sonsors ranging from 0-10k?, because somehow they impacted the AIN-values on neighboured lines! I was about to become mad. But now everything works very smooth! (sorry for the many edits :)

ahh one more thing: I have some sensors that range from 0k? - whatever k?. I got some problems with too low resistance values (it's fun frying resistors ;D ), so I used to take 1 k? as min instead and round the RL down a bit. It may not be the finest mathmatical thing, but at least I can use it :) Corrections welcome!

here you go: http://www.midibox.org/forum/index.php?topic=6506.0 I opened a new topic for these ;)

I have some 8cm softpots and two 360° pots by spectrasymbol. Also one very nice FSR-Sensor to measure finger-pressure (Conrad has these). The secret is, to get the load resistance value (impédance de charge, Lastwiderstand). The formula for this is: RL = ? (Rmin * Rmax) if(Rmin==0) the formula is: RL = ? Rmax (if you don't see ?, it's "square root", "Wurzel aus") Then you have to solder a load resistor (RL) between +5V and Signal (S); two of the pot pins have to be connected to Ground (GND) and Signal: +5V ----(RL)---+ | | S -----------+---(Sensor)----+ | | GND ---------------------------+ So here is an example from my FSR-sensor: - connect your multimeter to the middle pin and select either the left OR the right - measure the minimal available restistance value: Rmin = 1 k? - measure the maximal available resistance value: Rmax = 8 k? RL = ? ( 1*8 ) = 2.83 k? now, we don't have no 2.83 k?, so we simply choose the next lower 2.2 k? and connect this between +5V and Signal. +5V ----(2k2)---+ | | S ------------+----(1..8k)----+ | | GND ----------------------------+ The AIN now should read values between 0.5V (min) to 4V (max). To interpolate that value to 0..127 you can either connect an operation amplifier (that has to be described by someone else) or check out http://www.midibox.org/forum/index.php?topic=5270 where the interpolation by code is described. The idea is, to read the AIN value as 10bit value and calculate the rest. There's enough resolution to do this without loss of reading (although you loose processing units) ;) Michael Edit: small translation error

wow, the c function reference addition is also very welcome! Thanks TK!! :)

Yeah, I have to admit, that I've been lazy too (regarding my midibox) ;D But now I want to use it in July, so I'm speeding up... :o Right now, I'm "wasting" a bit of time in writing a small C-Debug Class (some kind of simulator), so I can debug my MIOS application within X-Code, that's really fine... I'll open a new C-Thread for the command-line simulator, once there's something to post ;) Cheers! Michael forgot to mention: I've also some spectra soft-pots connected (how could it be else? ;D ), if you have questions I can help... was a bit tricky to find out the right pull-up resistor values...

I'm currently cleaning up my sensorizer code in a seperate .c-file, together with lots of comments and an easy to tweak-setup. I have found some errors on the last pieces of code I posted, so if anybody wants to use it, there'll be an update shortly... Here are some features :o - en-/disable ains - en-/disable single processing features - sense threshold (don't sense below) - gate (incrementer, timeout, threshold) - pedal (and panic on release - optional) - expander (based on a sensor min value and a factor of 1/2/4 or 8 ) - scaler (scale resulting value to min (0) and max (127) << this will be quite cool if ya have two pots controlling min/max values!) - inverter (0-127 => 127-0) Also, with the new sensorizer, it is possible (but not needed) to avoid complex divisions! So it'll get a lot faster... and yes: it's already up and running 8) , but it'll still need some checking and improving ... Michael

have you connected something other than the distance sensor to check if the input is working correctly (adressed as Pin 0)? E.g. a 10k-Pot? Can you provide some more info? What's the code of your AIN_NotifyChange()? Are the values totally random or do they just flicker? All I can say from my point of view is that it should (at least with only one sensor) work as expected (besides not reaching exactly 0..5V and a lot of flickering)

Totally random values indicate that you have ungrounded AIN-pins together with a pre-existing application (like MB64)... right? You need to ground the unconnected pins or (if you're programming a custom app) you should only process pins that are connected (see the source examples). Does that help?

Yeah, Absynth has a PRG-CH Change Bug since 3.0. I start a small riot every month or so via NI forum and by emailing the support, but I'm waiting since a year now... so don't expect that anything happens (except that there will be a new release for 500$ with additional fresh bonus bugs... probably) grrr >:(